mummolo_data.csv Codebook

--NOTE 1: In order to try to filter respondents located outside the US, we collected respondent IP addresses (with permission from Vanderbilt University's IRB).  We also initially collected repsondents' approximate longitude and latitude.  We have removed this information from the replication data to protect respondent anonymity.

--NOTE 2: When read into R, only the first of the three rows in the .csv file containing headers from Qualtrics is retained, so the below variable names reflect those headers in the first row.


--NOTE 3: We have provided the complete raw Qualtrics responses (with an exception noted next).  Given the nature of fielding conjoint experiments with the ConjointSDT application, the .csv file contains a large number of columns that store attributes and levels but do not themselves have consistent meaning across respondents.  We explain the general purpose of those columns in the codebook but do not describe each of those columns individually.


[Variable descriptions provided in order left to right as they appear in .csv file]

QUALTRICS-SUPPLIED VARIABLES
StateDate--Date/time at which respondent started survey
EndDate--Date/time at which respondent completed survey
Status--Qualtrics indicator for how checks for multiple responses were monitored
Progress--Value indicating the % of the survey the respondent completed
Duration--Number of seconds it took the respondent to complete the survey
Finished--Logical indicating whether the respondent finished the survey
RecordedDate--Date/time at which responses were recorded
ResponseId--Randomly-generated alphanumeric value to identify each respondent
DistributionChannel--Factor indicating how respondents entered into the survey
UserLanguage--The language in which the respondent took the survey

PRE-TREATMENT SURVEY RESPONSE VARIABLES
Q2--Year respondent was born
Q4--Respondent gender (Male; Female; Other)
Q4_3_TEXT--If the respondent's gender is "other," the contents of the text box
Q6--Respondent educational attainment (Some high school, or less; High school graduate or GED; Some college, no 4-year degree; College graduate; Post-graduate degree)
Q7--Respondent race (American Indian or Alaska Native; Asian; Black or African-American; Native Hawaiian or other Pacific Islander; Other; White)
Q7_6_TEXT--If the respondent's race is "other," the contents of the text box
Q8--Respondent's identification as Hispanic (Yes; No)
Q9--Respondent's income (Less than $25,000; $25,000-$50,000;   $50,000-$75,000;  $75,000-$100,000; $100,000-$200,000; $200,000 or more)
Q11--Respondent's ideology (Very conservative, conservative, slightly conservative, moderate, slightly liberal, liberal, very liberal)
Q12--Respondent's initial partisan identification (Republican, Democrat, Independent, Other)
Q13--Whether respondents choosing Democrat in Q12 identify as strong or not strong Democrats (Strong Democrat; not a very strong Democrat)
Q14--Whether respondents choosing Republican in Q12 identify as strong or not strong Republicans (Strong Republican; not a very strong Republican)
Q15--Whether respondents choosing Indepedent in Q12 identify as leaners or true independents (Democratic, Republican, Neither)
Q16--Whether respondent has health insurance (Yes, No)
Q17--Whether respondent works in the health care industry (Yes, No)
Q18--Whether respondent is a student (Yes, I'm a full-time student; Yes, I'm a part-time student; No, I am not currently a student)
Q19--Whether respondent is trying to lose weight (Yes, No)
Q20--How many cigarettes the respondent smokes per week (I never smoke; One cigarette; One pack; Two to four packs; Five to ten packs; More than ten packs)

CONJOINT TASK/PROFILE RESPONSES
Except for column 29 (Q50), columns 28-52 contain respondents' answers to questions about the profiles featured in each conjoint task.  Respondents in the forced choice arm have their responses recorded in columns 28-40 (except column 29), while those in the abstention arm have their choices recorded in columns 41-52.  Q50 asked respondents to summarize the headline of the story they selected in the first task.

CONJOINT ATTRIBUTES AND LEVELS
Columns 53-124 provide the attributes and the corresponding levels presented in each conjoint task in each profile.  These attributes and levels were processed by Qualtrics as embedded data from a .php file designed by the Conjoint Survey Design Tool 2.0.  Headers with two numbers indicate the name of the attribute, and headers with three numbers indicate the level for the corresponding attribute.  For headers with two numbers, the first number indicates the task, and the second number indicates the order in which the attribute appeared in the conjoint table.  For headers with three numbers, the first number indicates the task, the second number indicates the profile within the task, and the third number indicates the order in which the attribute appeared in the conjoint table.  For instance, variable F-1-1 indicates the attribute that appeared first in the conjoint table in the first task, and variables F-1-1-1, F-1-2-1, and F-1-3-1 indicate the values that were provided for that attribute for the first, second, and third profiles, respectively.

MISCELLANEOUS
confirmation_code--Completion codes used by MTurkers to verify they completed the survey.
IP_block--Indicator for whether the respondent's IP address indicated that they should not be allowed to participate in the survey
IP_country--Country in which the respondent's IP address was located.
Remaining columns indicate display order; values FL_8_DO indicate whether the respondent was assigned to the forced choice (FL_15) or abstention (FL_16) arm, and the remaining columns indicate which task was repeated for the respondent's 13th task.